Implement _unique_axis_0 by with structured view

By viewing the array as a structured array that merely groups together
the other dimensions within the structure, `_unique_axis_0` is able to
call NumPy's `unique` function on the array keeping the type information
unchanged. Thus if `unique` is able to handle the specific type more
efficiently, we benefit from that speed up still.

Note: More work would be needed if we wanted to handle F-contiguous
arrays, but that is not needed for our use case.

dask_image/ndmeasure/_utils/_label.py

@@ -52,6 +52,14 @@ def relabel_blocks(block_labeled, new_labeling):
     return relabeled
 
 
+def _unique_axis_0(a):
+    """Find unique 0th axis in N-D array."""
+    dt = numpy.dtype([("values", a.dtype, a.shape[1:])])
+    av = a.view(dt)
+    r = numpy.unique(av)["values"]
+    return r
+
+
 def _across_block_label_grouping(face, structure):
     """Find a grouping of labels across block faces.